Tag
A post laying out how to get good at Generative UI, recommending Google's "LLMs are Effective UI Generators" paper, key frameworks like Vercel AI SDK, CopilotKit/AG-UI, A2UI and MCP Apps, plus practice projects and reference repos.
A step-by-step tutorial on building generative UI — where the AI decides the shape of the answer and the screen renders live cards, charts, maps and tables — using the open-source OpenUI framework on Claude Opus 5.5 or any other model via OpenRouter.
OpenUI 1.0 brings ChatGPT-style generative UI to any app or agent, working with both closed and open models by converting an app's component library into a system prompt and rendering LLM output in the product's own design system.
OpenUI 1.0 发布,这是一个开源框架,能让 AI 代理把文本回答转化为由产品自身组件构建的实时界面(卡片、表格、地图、表单),开发者选择组件、模型决定展示内容。
Shared an interactive 2.5D visualization page that breaks down how ChatGPT's intelligent UI works: the model doesn't write code directly — it outputs DIL language, which is compiled and streamed by the server, and rendered client-side using a pre-built component set.
提出了GenUI-Harness,一个多智能体框架,将信息检索的Tool Agent与生成前端代码界面的GUI Coder Agent配对,并引入UI-TauBench基准测试。4B模型经训练后Pass@3从9.33%提升至58.00%,超越Claude Opus 5(46.67%),生成的UI还将平均对话轮次从3.4次降至1.2次。
ChatGPT's newly launched 'Intelligent UI' (generative UI) feature was reverse engineered within 24 hours by a team using only publicly observable behavior, and they released an open-source framework, Open Intelligent UI, that lets developers recreate similar experiences with local LLMs via Ollama or LM Studio.
作者逆向工程了 OpenAI 新推出的 ChatGPT Intelligent UI(生成式 UI),解析其架构:模型输出一种名为 DIL 的中间语言(Markdown + JSX 风格标签 + JavaScript),由服务端编译为 JS 程序与 JSON 文档,再由客户端沙箱运行时渲染为原生组件。
OpenAI unveiled an 'Intelligent UI' update for ChatGPT that generates custom visual, interactive elements—such as diagrams, calculators, and seat maps—powered by its GPT-6 model, which is rolling out to paid and then free users.
Shopify launched Canvas, an AI-powered site-building tool that lets merchants create and customize their online stores by chatting with its AI agent, Sidekick, while seeing real-time renders of the actual store code.
Airbnb CEO Brian Chesky argues chatbots aren't the right interface for e-commerce and travel, saying consumer AI needs new deterministic/generative interfaces and that AI agents require their own operating layer. Airbnb is developing "multiplayer" AI and making its app more agent-friendly.
Chris Tate's experiment demonstrates an instant Generative UI system using json-render and jev, enabling real-time rendering of components with custom design systems.
OUI-1 is the world's first model for Generative UI, a finetuned DiffusionGemma that generates user interfaces in OpenUI Lang with speed and reliability on consumer hardware.
Runway introduces Solaris, an Interface World Model that generates interactive interfaces frame by frame in real time without code, outperforming frontier LLMs.
EvoGenUI-Bench introduces a benchmark for evaluating LLMs as multi-turn generative UI assistants, focusing on interface maintenance across 150 tasks with challenges in state propagation and external grounding.
The tweet highlights how improving AI model performance is driving more generative and dynamic user interfaces, mentioning new deployment filters in the Vercel dashboard.
Open Bot is an open source version of Grok Bot that works with any agent harness, featuring AI Coworkers, Generative UI, and full data recording capabilities.
Open Bot is an open-source Grok Bot designed for real companies, featuring AI Coworkers, Generative UI, and computer use capabilities that can run with any model and harness.
The paper proposes an Inverse Theory of Mind (IToM) pipeline that infers user beliefs, preferences, and decision-making traits from observed interactions, using LLM-driven counterfactual reasoning to synthesize structured user personas for adaptive content recommendation across modalities, including a VisionOS spatial banking app.
Diyi Yang announces a talk with Yanzhe Zhang about generative UI at ACL2026, discussing new interfaces for human-AI interaction, with Zhang presenting their GenUI work in a poster session.